Non-linear internal model control method based on Smith predictor and suitable for wall-hanging stove
Through the nonlinear internal mode control method based on Smith estimator, combined with feedback linearization and internal mode control, the problem that the gas wall-mounted furnace temperature control technology is difficult to meet the needs of complex working conditions is solved, and accurate and stable temperature control is achieved, adaptable System changes improve control accuracy and safety.
Patent Information
- Application Number
- CN202411013897.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-26
- Publication Date
- 2024-11-12
AI Technical Summary
The temperature control technology of existing gas wall-mounted furnaces is difficult to meet the high performance requirements of products with different capacity, power and structure under complex operating conditions, and traditional PID control algorithms are complex and difficult to achieve accurate constant temperature control.
Using a nonlinear internal mode control method based on Smith estimator, combined with feedback linearization control, control based on Smith estimator and internal mode control, it handles system nonlinearity, delay and external interference, and provides a complete control parameter tuning solution. , only one parameter to be adjusted is required.
It realizes the accuracy and stability of temperature control under complex operating conditions, offsets the impact of system nonlinearity and delay on the closed-loop system, adapts to system parameter changes, eliminates model uncertainty and external interference, and improves control accuracy and system safety. reliability.
Smart Images

Figure CN118938677A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of wall-mounted boilers, and in particular to a nonlinear internal model control method based on a Smith predictor suitable for wall-mounted boilers. Background Art
[0002] With the requirements of "carbon peak and carbon neutrality", gas supply facilities have become more popular in small and medium-sized cities, and the market demand for gas wall-mounted boilers has also expanded further. In recent years, my country's gas wall-mounted boiler market has gradually shifted from "popular consumption" to "comfort consumption", focusing on product upgrades from the perspective of user experience. In fact, users' requirements for product comfort run through the entire development process of gas wall-mounted boilers. Today, as people continue to pursue high-quality life, the requirements for comfort have been raised to an unprecedented level. Whether the product meets the comfort requirements even determines the market fate of gas wall-mounted boilers. In order to respond to the world's call for energy conservation and emission reduction and meet the increasingly diversified product needs of the market, Xiderui Group, as one of Europe's leading manufacturers of heating / hot water equipment, has improved the quality of its products in all aspects, including constant temperature control, heat exchange efficiency, combustion noise, insulation time, flue gas treatment, safety assurance, service life, product appearance, etc. Among them, the rapidity, stability and robustness of water temperature regulation are one of the decisive factors affecting user comfort and are also important indicators for product quality evaluation. Therefore, the temperature control technology of gas wall-mounted boilers is naturally placed at the top of the product's core technology.
[0003] Due to the increasing user needs, gas wall-mounted boilers have derived a variety of product models. There are large differences in size, power, structure, etc. between different products, and the factors affecting temperature control performance are more complicated. In order to achieve precise water temperature control, the traditional PID control algorithm improved based on engineering experience is becoming increasingly complex and bloated. Although the calculation amount of the control algorithm has doubled, it is still difficult to meet the high-performance requirements of the constant temperature control system of various products with different capacities, powers, and structures under different working conditions. It is urgent to design a more concise and efficient advanced control algorithm to replace the traditional engineering PID algorithm.
[0004] Based on the above situation, the present invention proposes a nonlinear internal model control method based on Smith predictor suitable for wall-mounted boilers, which can effectively solve the above problems. Summary of the invention
[0005] The purpose of the present invention is to provide a nonlinear internal model control method based on Smith Predictor suitable for wall-mounted boilers. The nonlinear internal model control method based on Smith Predictor suitable for wall-mounted boilers of the present invention embeds three control technologies: feedback linearization control, Smith Predictor-based control and internal model control, which are respectively used to deal with the influence of system nonlinearity, time delay, model parameter changes and external interference. This scheme can effectively ensure the accuracy and stability of temperature control of the wall-mounted boiler system under complex working conditions, and can not only reasonably offset the influence of system nonlinearity and time delay on the stability of the closed-loop system, but also flexibly adapt to the large changes in system parameters, fully eliminate the influence of model uncertainty and external interference on control accuracy, and finally ensure the safe and reliable operation of the entire temperature control system. In addition, the present invention provides a complete control parameter setting scheme with only one parameter to be adjusted.
[0006] The present invention is achieved through the following technical solutions:
[0007] The nonlinear internal model control method based on Smith Predictor applicable to wall-mounted boilers includes:
[0008] Step S1: Summarize the heat transfer process of the dual-purpose gas wall-mounted boiler system in DHW mode and establish a nonlinear model;
[0009] Step S2: using feedback linearization method to convert the nonlinear model into a linear model to deal with the influence of actuator nonlinearity on the stability of the closed-loop system;
[0010] Step S3: designing a Smith Predictor-based control structure for the linear model to handle the effect of time delay on the closed-loop system;
[0011] Step S4: deriving a parameter setting method based on the internal model control principle;
[0012] Step S5: performing stability analysis on the closed-loop system;
[0013] Step S6: Based on the test data, obtain the identification results of each part of the linear model;
[0014] Step S7: Perform a cold start test, a water outage test and a water flow change test on the water temperature regulating system.
[0015] The nonlinear internal model control method based on Smith Predictor suitable for wall-mounted boilers of the present invention embeds three control technologies: feedback linearization control, Smith Predictor-based control and internal model control, which are respectively used to deal with the influence of system nonlinearity, time delay, model parameter changes and external interference. This scheme can effectively ensure the accuracy and stability of temperature control of the wall-mounted boiler system under complex working conditions, can reasonably offset the influence of system nonlinearity and time delay on the stability of the closed-loop system, and can flexibly adapt to the large changes in system parameters, fully eliminate the influence of model uncertainty and external interference on control accuracy, and finally ensure the safe and reliable operation of the entire temperature control system. In addition, the present invention provides a complete control parameter setting scheme, with only one parameter to be adjusted.
[0016] Preferably, the specific steps of step S1 include:
[0017] Step S11: Ignoring the instantaneous dynamics of the gas valve adjustment and the combustion process, and ignoring the slight deviation of the gas valve return stroke, the thermal power Q b (t) and gas pressure P gv The relationship between (t) satisfies:
[0018]
[0019] Among them, ζ can be regarded as an approximately constant coefficient, which is related to the calorific value of the gas, the nozzle coefficient, the nozzle diameter and the relative density of the gas;
[0020] Step S12: Gas pressure P gv (t) and driving current i gv The nonlinear relationship between (t) can be approximately expressed as:
[0021]
[0022] Among them, i min with i max are the minimum and maximum operating currents of the gas valve, P min With P max are the minimum and maximum gas pressures that the gas valve can supply, i op and i st They correspond to the actual opening current of the gas valve and the current when it reaches saturation, respectively. c1, c2, and c3 are constant coefficients.
[0023] Step S13: Ignore the small energy loss during the internal water circulation process, that is, assume that T ω (t) = T ω (t-τ h -τ ω ), the energy conservation equation of the water heating system can be summarized as:
[0024]
[0025] Where η is the heat conversion efficiency of the combustion process, ρ (kg / m 3 ) is the density of water, m(kg) is the mass of the domestic water capacity in the secondary plate exchange (with W f linear relationship), c e (J / kg·℃) is the specific heat capacity of water;
[0026] Step S14: Set x(t)=T o (t)-T c Considered as system state, i gv (t) is regarded as the active input (i.e., control signal), and T c Considered as passive input (i.e. load), the nonlinear model of the dual-purpose wall-mounted boiler system in DHW mode can be derived by combining equations (1) to (3):
[0027]
[0028] Where τ = τ b +τ h and
[0029] Please note that for measuring T o The temperature sensor is located at the outlet of the secondary plate heat exchanger, τ h With water flow rate W f In a linear relationship, T c This term can be considered as the initial value of y(t).
[0030] Preferably, the specific steps of step S2 include:
[0031] Step S21: Due to the existence of f(·), in order to linearize the nonlinear model, an intermediate control signal v is introduced here m , which is defined as v m (t) = f[i gv (t)];v m The physical meaning of is the square root value of gas pressure, that is In addition, define the model parameter α = ρW f / m and β = ηζ / mc e , then (4) can be rewritten as:
[0032]
[0033] If the water flow rate W f unchanged, the parameters α and β are constants, so at the water flow rate W f Under the condition of no change, the nonlinear model is transformed into a linear model through the above feedback linearization process;
[0034] Step S22: After the control signal designed for the linear model is determined, the physical control signal acting on the real nonlinear model, that is, the gas valve current, should be calculated by the following formula:
[0035] i gv (t) = f inv [v m (t)]=f -1 [v m (t)] (6)
[0036] Among them, f(·) and f inv (·) Obtained through system identification.
[0037] Preferably, the specific steps of step S3 are:
[0038] Step S31: Perform Laplace transform on the linear model:
[0039]
[0040] Among them, X(s), Y(s) and V m (s) are x(t), y(t) and v m (t) in the s-domain form;
[0041] Step S32: Based on this, under zero initial conditions, the transfer function of the linear model can be extracted as:
[0042]
[0043] Among them, K m =β / α=η / (ζρc e W f ) and H m =1 / α=m / (ρW f );
[0044] Step S33: G m (s) is decomposed into pure delay part e -τs and the delay-free part G m0 :
[0045]
[0046] Step S34: Based on the models (2), (8) and the delay-free model (9) obtained by system identification, a control structure based on the Smith Predictor is designed. The transfer function of the Smith Predictor can be summarized as:
[0047] G sp (s) = G m0 (s)-G m(s) = G m0 (s)(1-e -τs ) (10)
[0048] Among them, the model parameters (K m , H m and time delay τ) are based on the system identification results and the water flow velocity measurement value W f Calculated in real time;
[0049] The feedback controller C(s) is designed as follows:
[0050]
[0051] Among them, k P and k I are the proportional gain and integral gain of the controller respectively.
[0052] Preferably, the specific steps of step S4 are:
[0053] Step S41: The equivalent condition of the Smith Predictor-based control structure and the IMC structure is to design C(s) so that the following equation holds:
[0054]
[0055] Step S42: It can be deduced that in order to make the control structure based on the Smith Predictor have the equivalent characteristics of the classic IMC structure, that is, the model adaptation and disturbance suppression capabilities, the controller C(s) needs to satisfy:
[0056]
[0057] Step S43: Design the equivalent internal model controller C for the first-order time delay model (8) IMC (s), model (8) can be factorized into G m (s) = G m0 (s)e -τs , where G m0 is a first-order minimum phase system; ideally, the internal model controller C IMC (s) should be G m0 (s) is the inverse of the closed-loop system. p (s) = G m (s) satisfies Y(s) / T set (s) = 1, thus achieving perfect tracking of the reference signal; however, due to Is not true, that is, the controller It is physically impossible to realize;
[0058] Therefore, here in C IMC(s) further introduces a first-order filter Q(s) = 1 / (λs+1), where λ is the filter time constant, thus converting the controller C IMC (s) is converted into a half-true model; In summary, the controller C IMC The final form of (s) is:
[0059]
[0060] Step S44: Combining (13) and (14), the specific expected form of the controller C(s) is:
[0061]
[0062] Step S45: Combining the original form (11) and the desired form (15) of the controller C(s), derive the proportionality coefficient k P and the integration coefficient k I Self-tuning scheme:
[0063]
[0064] Among them, the model parameter K m , H m and τ should be based on the system identification results and the water flow rate measurement value W f Real-time calculation, will change with the water flow rate W f changes with fluctuations.
[0065] Preferably, the specific steps of step S5 are:
[0066] Step S51: Since the transfer function of the control structure based on the Smith Predictor is as follows:
[0067]
[0068] Therefore, the closed-loop system satisfies:
[0069]
[0070] Where D(s) represents the form of d(t) in the s domain;
[0071] Step S52: Set G p The delay-free part in (s) is G p0 (s)(i.e. G m0 (s) corresponds to the real model), the real delay of the system is τ p (i.e. the true value of the total delay τ), substituting (15) and (17) into (18), we can obtain:
[0072]
[0073] It can be seen from formulas (19) and (20) that if G m ≈G p , then the closed-loop system can be decomposed into a delay-free part CG p0 / [1+C(G m0 -G m +G p )] and a pure delay part
[0074] Because when So from (19) and (20) we can infer that:
[0075]
[0076] That is, when s→0, the output Y(s) will reach T set (s), and the effect of interference D(s) on output Y(s) is 0;
[0077] Step S53: Set the system nonlinearity f inv The model errors of and f are regarded as part of the model uncertainty and external disturbance. Then, combined with the final value theorem, it can be seen from (19) that if the input signal T set (t) and d(t) are bounded and stable even in the presence of time delays, model uncertainties, external disturbances, and f The parameter changes caused by the change, as t→∞, the system output y(t)=T o (t) will still accurately reach the set value T set , and the influence of interference on the output will be completely eliminated.
[0078] Preferably, the specific steps of step S6 are:
[0079] Step S61: When the water flow rate remains unchanged, v(t) = P b (t) = f[i gv (t)] to T o The transfer function of (t) can be summarized as:
[0080] Among them, K m =β / α=η / (ζρc e W f ) and H m =1 / α=m / (ρW f );
[0081] Step S62: Based on the test data, use the Curve Fitting toolbox of Matlab to perform fitting to obtain the identification results of each part of the system model:
[0082]
[0083] τ≈2.04s.
[0084] Compared with the prior art, the present invention has the following advantages and beneficial effects:
[0085] 1. A multi-condition modeling method for the domestic hot water mode of a dual-purpose wall-mounted boiler system is provided. This method comprehensively considers complex factors such as actuator nonlinearity, changes in system parameters with water flow, and time delay.
[0086] 2. Identify system model parameters to effectively improve model accuracy and enhance control quality.
[0087] 3. Feedback linearization control technology is embedded to effectively deal with the impact of actuator nonlinearity on temperature control stability.
[0088] 4. The Smith predictor-based control technology is embedded to effectively handle the impact of large time delays on temperature control stability.
[0089] 5. The internal model control technology is embedded, which can flexibly adapt to the large changes in system parameters, fully eliminate the influence of model uncertainty and external interference on control accuracy, and effectively ensure the stability and accuracy of temperature control of the system under harsh working conditions such as cold and hot start-up and large changes in water flow rate.
[0090] 6. A complete parameter setting scheme is designed based on the internal model control principle. There is only one parameter to be adjusted, which greatly reduces the number of control parameters, reduces the workload of front-line debugging personnel, and improves the intelligence of the product.
[0091] 7. The stability, control accuracy and anti-interference performance of the designed closed-loop system were rigorously analyzed. The entire control scheme effectively improved the key indicators of the gas wall-mounted boiler system, such as the temperature convergence speed and steady-state control accuracy. BRIEF DESCRIPTION OF THE DRAWINGS
[0092] Figure 1 It is a schematic flow chart of the heat transfer process of the dual-purpose gas wall-mounted boiler in the DHW mode of the present invention;
[0093] Figure 2 This is a schematic flow chart of the control based on the Smith Predictor of the present invention;
[0094] Figure 3 It is a schematic flow chart of the equivalent between the control structure based on Smith Predictor and the IMC structure of the present invention;
[0095] Figure 4 This is a comparison diagram of water temperature curves with a target temperature of 42° C. in the cold start test of the present invention;
[0096] Figure 5 This is a comparison diagram of water temperature curves with a target temperature of 60° C. in the cold start test of the present invention;
[0097] Figure 6 This is a comparison diagram of water temperature curves with a target temperature of 42° C. in the water flow change test of the present invention;
[0098] Figure 7 This is a comparison diagram of water temperature curves with a target temperature of 60° C. in the water flow change test of the present invention. DETAILED DESCRIPTION
[0099] In order to enable those skilled in the art to better understand the technical solution of the present invention, the preferred implementation scheme of the present invention is described below in conjunction with specific embodiments. However, it should be understood that the drawings are only used for exemplary description and cannot be understood as a limitation on this patent. In order to better illustrate this embodiment, some parts of the drawings may be omitted, enlarged or reduced, and do not represent the size of the actual product. For those skilled in the art, it is understandable that some well-known structures and their descriptions in the drawings may be omitted. The positional relationship described in the drawings is only for exemplary description and cannot be understood as a limitation on this patent.
[0100] Embodiment 1:
[0101] like Figures 1 to 7 As shown, the present invention provides a nonlinear internal model control method based on Smith predictor applicable to a wall-mounted boiler, comprising:
[0102] Step S1: The main components of the wall-mounted boiler include the main heat exchanger, secondary plate heat exchanger, burner, water pump, gas valve, various sensors and pipelines. When the system is in DHW mode, the three-way valve will close the heating hot water pipe. Then, the gas valve is set by a given current l gv (mA) drive to allow a certain pressure P gv (mbar) of gas passes through. Through the combustion reaction, the gas flowing through the burner converts its chemical energy into thermal energy, heating the air in the chamber. b After a few seconds of combustion, heating and air flow, the hot air reaches the main heat exchanger and heats the internal circulating water to a temperature of T h (℃) and output. In other words, the heat energy released by the burner is calculated according to the heat power Q b (J / min) is transferred to the internal circulating water. The internal circulating water flow rate is W f (L / min). After τ h seconds of flow time, the heated internal circulating water reaches the secondary plate heat exchanger, and the temperature is T c (℃) of cold water inlet is heated to outlet water temperature T o(℃), thus providing the family with hot water at a constant temperature. At the same time, the internal circulating water temperature drops to
[0103] T ω (℃), after τ ω After a flow time of seconds, it will return to the main heat exchanger. At this point, the entire heat transfer process is over. Summarize the heat transfer process of the dual-purpose gas wall-mounted boiler system in DHW mode and establish a nonlinear model; Figure 1 The heat transfer process in the domestic hot water mode (DHW) is summarized. It can be seen that the system consists of two subsystems, namely the combustion system consisting of the gas valve and the burner, and the water heating system consisting of the heat exchanger and the pipe. gv , domestic water can be heated to the set temperature T set (℃).
[0104] The control objective of the present invention is to adjust the gas valve current so that the outlet temperature reaches the set temperature, that is, by adjusting the control signal i gv Make the system output y(t) = T o (t) reaches the set value T set . Considering the adverse effects of actuator nonlinearity, system parameter changes, model uncertainty, external disturbances and system delay on control performance, in order to reliably achieve the goal. Step S2: Use feedback linearization method to convert the nonlinear model into a linear model to deal with the influence of actuator nonlinearity on the stability of the closed-loop system;
[0105] Step S3: designing a Smith Predictor-based control structure for the linear model to handle the effect of time delay on the closed-loop system;
[0106] Step S4: deriving a parameter setting method based on the internal model control principle;
[0107] Step S5: performing stability analysis on the closed-loop system;
[0108] Step S6: Based on the test data, obtain the identification results of each part of the linear model; the test data refers to the domestic hot water test performed on the wall-mounted boiler during modeling, including cold start, etc., see the curve named "BDR" in the test data of this patent for details.
[0109] Step S7: Perform a cold start test, a water outage test and a water flow change test on the water temperature regulating system.
[0110] Embodiment 2:
[0111] like Figures 1 to 7 As shown, the present invention provides a nonlinear internal model control method based on Smith predictor applicable to a wall-mounted boiler, comprising:
[0112] Step S1: Summarize the heat transfer process of the dual-purpose gas wall-mounted boiler system in DHW mode and establish a nonlinear model;
[0113] Step S2: using feedback linearization method to convert the nonlinear model into a linear model to deal with the influence of actuator nonlinearity on the stability of the closed-loop system;
[0114] Step S3: designing a Smith Predictor-based control structure for the linear model to handle the effect of time delay on the closed-loop system;
[0115] Step S4: deriving a parameter setting method based on the internal model control principle;
[0116] Step S5: performing stability analysis on the closed-loop system;
[0117] Step S6: Based on the test data, obtain the identification results of each part of the linear model;
[0118] Step S7: Perform a cold start test, a water outage test and a water flow change test on the water temperature regulating system.
[0119] The specific steps of step S1 include:
[0120] Step S11: Ignoring the instantaneous dynamics of the gas valve adjustment and the combustion process, and ignoring the slight deviation of the gas valve return stroke, the thermal power Q b (t) and gas pressure P gv The relationship between (t) satisfies:
[0121]
[0122] Among them, ζ can be regarded as an approximately constant coefficient, which is related to the calorific value of the gas, the nozzle coefficient, the nozzle diameter and the relative density of the gas;
[0123] Step S12: Gas pressure P gv (t) and driving current i gv The nonlinear relationship between (t) can be approximately expressed as:
[0124]
[0125] Among them, i min with i max are the minimum and maximum operating currents of the gas valve, P min With P max are the minimum and maximum gas pressures that the gas valve can supply, i op and i st They correspond to the actual opening current of the gas valve and the current when it reaches saturation, respectively. c1, c2, and c3 are constant coefficients.
[0126] Step S13: Ignore the small energy loss during the internal water circulation process, that is, assume that T ω (t) = T ω (t-τ h -τ ω ), the energy conservation equation of the water heating system can be summarized as:
[0127]
[0128] Where η is the heat conversion efficiency of the combustion process, ρ (kg / m 3 ) is the density of water, m(kg) is the mass of the domestic water capacity in the secondary plate exchange (with W f linear relationship), c e (J / kg·℃) is the specific heat capacity of water;
[0129] Step S14: Set x(t)=T o (t)-T c Considered as system state, i gv (t) is regarded as the active input (i.e., control signal), and T c Considered as passive input (i.e. load), the nonlinear model of the dual-purpose wall-mounted boiler system in DHW mode can be derived by combining equations (1) to (3):
[0130]
[0131] Where τ = τ b +τ h and
[0132] Please note that for measuring T o The temperature sensor is located at the outlet of the secondary plate heat exchanger, τ h With water flow rate W f In a linear relationship, T c This term can be considered as the initial value of y(t).
[0133] Furthermore, in another embodiment, the specific steps of step S2 include:
[0134] Step S21: Due to the existence of f(·), in order to linearize the nonlinear model, an intermediate control signal v is introduced here m , which is defined as v m (t) = f[i gv (t)];v m The physical meaning of is the square root value of gas pressure, that is In addition, define the model parameter α = ρW f / m and β = ηζ / mce , then (4) can be rewritten as:
[0135]
[0136] If the water flow rate W f unchanged, the parameters α and β are constants, so at the water flow rate W f Under the condition of no change, the nonlinear model is transformed into a linear model through the above feedback linearization process;
[0137] On this basis, the control target is transformed into adjusting the intermediate control signal v m Output T as operating system o Tracking T set . Note that f(·) and f inv (·) is a bounded monotone non-decreasing function. If the linear model is stable, then the nonlinear model is also stable. In other words, the feedback linearization process does not affect the stability of the closed-loop system.
[0138] Step S22: After the control signal designed for the linear model is determined, the physical control signal acting on the real nonlinear model, that is, the gas valve current, should be calculated by the following formula:
[0139] i gv (t) = f inv [v m (t)]=f -1 [v m (t)] (6)
[0140] Among them, f(·) and f inv (·) Obtained through system identification.
[0141] Further, in another embodiment, the sampling time of the dual-purpose wall-mounted boiler is only 0.1 seconds. f is small enough, the total delay τ in model (10) can exceed 4 seconds; if W f If τ is large enough, then τ may be less than 0.5 seconds. Therefore, relative to the sampling time of the wall-mounted boiler, the total delay of the system may not only be very large, but also may change dramatically, which is likely to change the stability of the closed-loop system. To solve this problem, the present invention designs a control structure based on the Smith Predictor for the linear model after linearization, and the specific steps are as follows:
[0142] Step S31: Perform Laplace transform on the linear model:
[0143]
[0144] Among them, X(s), Y(s) and V m (s) are x(t), y(t) and vm (t) in the s-domain form;
[0145] Step S32: Based on this, under zero initial conditions, the transfer function of the linear model can be extracted as:
[0146]
[0147] Among them, K m =β / α=η / (ζρc e W f ) and H m =1 / α=m / (ρW f );
[0148] Step S33: G m (s) is decomposed into pure delay part e -τs and the delay-free part G m0 :
[0149]
[0150] Step S34: Based on the models (2), (8) and the delay-free model (9) obtained by system identification, a control structure based on the Smith Predictor is designed, such as Figure 2 As shown, Figure 2 e=T set -T o , G p (s) represents G m (s) is the true model of the dynamics described, F -1 and F stands for f inv and the s-domain form of f;
[0151] Figure 2 The transfer function of the Smith Predictor can be summarized as:
[0152] G sp (s) = G m0 (s)-G m (s) = G m0 (s)(1-e -τs ) (10)
[0153] Among them, the model parameters (K m , H m and time delay τ) are based on the system identification results and the water flow velocity measurement value W f Calculated in real time;
[0154] The feedback controller C(s) is designed as follows:
[0155]
[0156] Among them, k P and k I are the proportional gain and integral gain of the controller respectively. Note that k in equation (17) P and k I The IMC-based parameter setting procedure (see the next section for details) is used according to the measured water flow rate W f Real-time adjustment. f When k remains unchanged, P and k I It is a definite constant.
[0157] Furthermore, in another embodiment, the specific steps of step S4 are:
[0158] Since the model parameter K m , H m and τ will change with the water flow rate W f Therefore, we hope that the controller gain k P and k, can self-tune to adapt to the changes in model parameters, thereby making the control performance of the closed-loop system more sensitive to the water flow rate W f In other words, it is hoped that the identified process model (whose model parameters should be based on the measured W f Real-time adjustment) is embedded into the controller to improve the control performance, which is consistent with the core idea of the IMC method.
[0159] To be in Figure 2 The key to realizing this idea in the control structure shown in Figure 1 is to establish the equivalence between the control structure based on the Smith predictor and the IMC structure, and then derive the parameter tuning scheme of the controller C(s) on this basis. Figure 2 The control structure based on Smith predictor shown in the figure can be restructured into Figure 3 The structure shown. m0 (s), F -1 The closed-loop system composed of and F is regarded as a submodule. If the controller C(s) is designed so that the module is equivalent to an internal model controller C IMC (s), then Figure 3 The Smith Predictor based control structure shown will be converted to a classical IMC structure.
[0160] Step S41: The equivalent condition of the Smith Predictor-based control structure and the IMC structure is to design C(s) so that the following equation holds:
[0161]
[0162] Step S42: It can be deduced that in order to make the control structure based on the Smith Predictor have the equivalent characteristics of the classic IMC structure, that is, the model adaptation and disturbance suppression capabilities, the controller C(s) needs to satisfy:
[0163]
[0164] Step S43: Design the equivalent internal model controller C for the first-order time delay model (8) IMC (s), model (8) can be factorized into G m (s) = G m0 (s)e -τs , where G m0 is a first-order minimum phase system; ideally, the internal model controller C IMC (s) should be G m0 (s) is the inverse of the closed-loop system. p (s) = G m (s) satisfies Y(s) / T set (s) = 1, thus achieving perfect tracking of the reference signal; however, due to Is not true, that is, the controller It is physically impossible to realize;
[0165] Therefore, here in C IMC (s) further introduces a first-order filter Q(s) = 1 / (λs+1), where λ is the filter time constant, thus converting the controller C IMC (s) is converted into a half-true model; In summary, the controller C IMC The final form of (s) is:
[0166]
[0167] Step S44: Combining (13) and (14), the specific expected form of the controller C(s) is:
[0168]
[0169] Step S45: Combining the original form (11) and the desired form (15) of the controller C(s), derive the proportionality coefficient k P and the integration coefficient k I Self-tuning scheme:
[0170]
[0171] Among them, the model parameter K m , H m and τ should be based on the system identification results and the water flow rate measurement value W f Real-time calculation, will change with the water flow rate W fchanges with fluctuations.
[0172] At this point, the design of the parameter tuning program based on IMC has been completed. It can be seen that using the parameter tuning program based on IMC, only one parameter needs to be adjusted, namely the time constant λ of the filter, which greatly simplifies the parameter tuning process of the controller C(s). In addition, the use of the parameter tuning program based on IMC can make Figure 2 The control structure shown is equivalent to the classical IMC structure, which makes the designed control system have the advantages of the IMC scheme, namely model adaptation and disturbance suppression capabilities.
[0173] In summary, the design of the entire control scheme combines three control techniques: feedback linearization control, Smith predictor-based control, and internal model control. Although the embedding process of these control techniques is introduced separately, the final implementation form of the scheme is a whole (such as Figure 2 shown).
[0174] Further, in another embodiment, the specific steps of step S5 are:
[0175] Step S51: Since the transfer function of the control structure based on the Smith Predictor is as follows:
[0176]
[0177] Therefore, the closed-loop system satisfies:
[0178]
[0179] Where D(s) represents the form of d(t) in the s domain;
[0180] Step S52: Set G p The delay-free part in (s) is G p0 (s)(i.e. G m0 (s) corresponds to the real model), the real delay of the system is τ p (i.e. the true value of the total delay τ), substituting (15) and (17) into (18), we can obtain:
[0181]
[0182] It can be seen from formulas (19) and (20) that if G m ≈G p , then the closed-loop system can be decomposed into a delay-free part CG p0 / [1+C(G m0 -G m +G p )] and a pure delay part Therefore, as long as G p and Gm If the difference between is small enough so that the characteristics of the delay-free part do not change dramatically, the change of system delay will not affect the stability of the closed-loop system, but will only change the tracking speed of the reference signal. This explains the working mechanism of the control structure based on the Smith Predictor to eliminate the influence of system delay.
[0183] So from (19) and (20) we can infer that:
[0184]
[0185] That is, when s→0, the output Y(s) will reach T set (s), and the effect of interference D(s) on output Y(s) is 0;
[0186] Step S53: Set the system nonlinearity f inv The model errors of and f are regarded as part of the model uncertainty and external disturbance. Then, combined with the final value theorem, it can be seen from (19) that if the input signal T set (t) and d(t) are bounded and stable even in the presence of time delays, model uncertainties, external disturbances, and f The parameter changes caused by the change, as t→∞, the system output y(t)=T o (t) will still accurately reach the set value T set , and the influence of interference on the output will be completely eliminated.
[0187] In practical applications, considering the hardware limitations and energy input limitations, the input signal T set The boundedness and stability conditions of (t) and d(t) are naturally satisfied. Therefore, based on the above analysis, the proposed control scheme can ensure the stability of the system and achieve the required control accuracy in industrial applications.
[0188] At this point, the stability analysis of the closed-loop system has been completed.
[0189] Furthermore, in another embodiment, the specific steps of step S6 are:
[0190] Step S61: When the water flow rate remains unchanged, v(t) = P b (t) = f[i gv (t)] to T o The transfer function of (t) can be summarized as:
[0191] Among them, K m =β / α=η / (ζρc e W f ) and H m=1 / α=m / (ρW f );
[0192] Step S62: Based on the test data, use the Curve Fitting toolbox of Matlab to perform fitting to obtain the identification results of each part of the system model:
[0193]
[0194] τ≈2.04s.
[0195] Cold start test:
[0196] 42℃ cold start test
[0197] The target temperature is 42℃, and the cold start curve of the gas water heater system is as follows: Figure 4 As shown:
[0198] based on Figure 4 Curve, the gas water heater system meets the index data in Table 1 below:
[0199] Table 1 Comparison of cold start indicators at 42℃
[0200]
[0201] Due to the limitation of the control amount, when the water inlet is 4 L / min and the target temperature is 42°C, both types of controllers cannot converge to the target value, but the solution proposed in the present invention can reach a stable value faster.
[0202] 60℃ cold start test
[0203] The target temperature is 60℃, and the cold start curve of the gas water heater system is as follows: Figure 5 As shown:
[0204] based on Figure 5 Curve, the gas water heater system meets the index data in Table 2 below:
[0205] Table 2 Comparison of cold start indicators at 60℃
[0206]
[0207] Compared with the original controller, when the water inlet is 4 L / min and the target temperature is 60° C., the control scheme proposed by the present invention has faster system dynamics and smaller overshoot.
[0208] Water cut test (hot start)
[0209] The results of the hot start test of the gas water heater are shown in Table 3 below:
[0210] Table 3 Hot start test
[0211]
[0212]
[0213] It can be seen from Table 3 that under hot start conditions, the controller proposed in the present invention has a smaller overshoot.
[0214] Water flow rate change test
[0215] In this part of the test, after setting the reference temperature, the water inlet flow rate is changed according to: 4-8-5-7-3-6 (L / min), and the water outlet temperature and gas pressure (GMP) are recorded.
[0216] The test results of 42℃ water flow rate change are as follows Figure 6 As shown in the figure, when the water inlet flow rate is 3-5 (L / min), the gas pressure has reached the minimum value and the control amount is in a saturated state. Figure 6 Curve, the gas water heater system meets the index data in Table 4 below:
[0217] Table 4 Constant temperature 42℃ curve index
[0218]
[0219] In the water flow change test at a target temperature of 42°C, the controller provided by the present invention has a faster stabilization speed, a smaller overshoot, and a smoother gas pressure.
[0220] The test results of 60℃ water flow rate change are as follows Figure 7 As shown in the figure, when the water inlet flow rate is 3 (L / min), the gas pressure has reached the minimum value and the control amount is in a saturated state. Figure 7 Curve, the gas water heater system meets the index data in Table 5 below:
[0221] Table 5 Constant temperature 60℃ indicators
[0222]
[0223]
[0224] In the water flow change test at a target temperature of 60°C, the controller provided by the present invention has a faster stabilization speed, a smaller overshoot, and a smoother gas pressure.
[0225] According to the description and drawings of the present invention, those skilled in the art can easily manufacture or use the Smith Predictor-based nonlinear internal model control method suitable for a wall-mounted boiler of the present invention, and can produce the positive effects described in the present invention.
[0226] Unless otherwise specified, in the present invention, the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise", "axial", "radial", "circumferential" and the like indicating orientation or positional relationships are based on the orientation or positional relationships shown in the drawings, which are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operate in a specific orientation. Therefore, the terms describing the orientation or positional relationship in the present invention are only used for illustrative purposes and cannot be understood as limitations on this patent. For ordinary technicians in this field, the specific meanings of the above terms can be understood in conjunction with the drawings and according to specific circumstances.
[0227] Unless otherwise clearly specified and limited, in the present invention, the terms "disposed", "connected" and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be directly connected, or indirectly connected through an intermediate medium, or it can be the internal communication of two elements. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.
[0228] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any form. Any simple modification or equivalent change made to the above embodiment based on the technical essence of the present invention shall fall within the protection scope of the present invention.
Claims
1. A nonlinear internal model control method based on Smith predictor applicable to wall-mounted boilers, characterized in that: include: Step S1: Summarize the heat transfer process of the dual-purpose gas wall-mounted boiler system in DHW mode and establish a nonlinear model; Step S2: Using feedback linearization method to convert the nonlinear model into a linear model to deal with the influence of actuator nonlinearity on the stability of the closed-loop system; Step S3: Designing a Smith Predictor-based control structure for the linear model to handle the effect of time delay on the closed-loop system; Step S4: deriving a parameter tuning method based on the internal model control principle; Step S5: performing stability analysis on the closed-loop system; Step S6: Based on the test data, obtain the identification results of each part of the linear model; Step S7: performing a cold start test, a water outage test, and a water flow change test on the water temperature regulating system.
2. The nonlinear internal model control method based on Smith Predictor for a wall-mounted boiler according to claim 1, characterized in that: The specific steps of step S1 include: Step S11: Ignoring the instantaneous dynamics of the gas valve adjustment and the combustion process, and ignoring the slight deviation of the gas valve return stroke, the thermal power Q b (t) and gas pressure P gv The relationship between (t) satisfies: Among them, ζ can be regarded as an approximately constant coefficient, which is related to the calorific value of the gas, the nozzle coefficient, the nozzle diameter and the relative density of the gas; Step S12: Gas pressure P gv (t) and driving current i gv The nonlinear relationship between (t) can be approximately expressed as: Among them, i min with i max are the minimum and maximum operating currents of the gas valve, P min With P max are the minimum and maximum gas pressures that the gas valve can supply, i op and i st They correspond to the actual opening current of the gas valve and the current when it reaches saturation, respectively. c1, c2, and c3 are constant coefficients. Step S13: Ignore the small energy loss during the internal water circulation process, that is, assume that T ω (t) = T ω (t-τ h -τ ω ), the energy conservation equation of the water heating system can be summarized as: Where η is the heat conversion efficiency of the combustion process, ρ (kg / m 3 ) is the density of water, m (kg) is the mass of the living water capacity in the secondary plate exchange (with W f linear relationship), c e (J / kg·℃) is the specific heat capacity of water; Step S14: Set x(t)=T o (t)-T c Considered as system state, i gv (t) is considered as active input (i.e., control signal), and T c Considered as a passive input (i.e. load), the nonlinear model of the dual-purpose wall-mounted boiler system in DHW mode can be derived by combining equations (1) to (3): where τ = τ b + τ h and Please note that for measuring T o The temperature sensor is located at the outlet of the secondary plate heat exchanger, τ h With water flow rate W f In a linear relationship, T c This term can be considered as the initial value of y(t).
3. The nonlinear internal model control method based on Smith Predictor for a wall-mounted boiler according to claim 2, characterized in that: The specific steps of step S2 include: Step S21: Due to the existence of f(·), in order to linearize the nonlinear model, an intermediate control signal v is introduced here m , which is defined as v m (t) = f[i gv (t)];v m The physical meaning of is the square root of gas pressure, that is In addition, define the model parameter α=ρW f / m and β=ηζ / mc e , (4) can be rewritten as: If the water flow rate W f If the water flow rate W is constant, the parameters α and β are constants. f Under the condition of no change, the nonlinear model is transformed into a linear model through the above feedback linearization process; Step S22: After the control signal designed for the linear model is determined, the physical control signal acting on the real nonlinear model, i.e., the gas valve current, should be calculated using the following formula: i gv (t)=f inv [v m (t)]=f -1 [v m (t)] (6) Among them, f(·) and f inv (·) Obtained through system identification.
4. The nonlinear internal model control method based on Smith Predictor for a wall-mounted boiler according to claim 3, characterized in that: The specific steps of step S3 are: Step S31: Perform Laplace transform on the linear model: Among them, X(s), Y(s) and V m (s) are x(t), y(t) and v m The s-domain form of (t); Step S32: Based on this, under zero initial conditions, the transfer function of the linear model can be extracted as: Among them, K m = β / α = η / (ζρc e W f ) and H m = 1 / α = m / (ρW f ); Step S33: G m (s) is decomposed into the pure delay part e -τs and the delay-free part G m0 : Step S34: Based on the models (2), (8) and the delay-free model (9) obtained by system identification, a control structure based on the Smith Predictor is designed. The transfer function of the Smith Predictor can be summarized as: G sp (s)=G m0 (s)-G m (s)=G m0 (s)(1-e -τs ) (10) Among them, the model parameters (K m , H m and time delay τ) are based on the system identification results and the water flow velocity measurement value W f Calculated in real time; The feedback controller C(s) is designed as follows: Among them, k P and k I are the proportional gain and integral gain of the controller respectively.
5. The nonlinear internal model control method based on Smith Predictor for a wall-mounted boiler according to claim 4, characterized in that: The specific steps of step S4 are: Step S41: The equivalent condition of the Smith Predictor-based control structure and the IMC structure is to design C(s) so that the following equation holds: Step S42: It can be deduced that in order for the Smith Predictor-based control structure to have the equivalent characteristics of the classic IMC structure, namely, model adaptability and disturbance rejection capabilities, the controller C(s) must satisfy: Step S43: Design the equivalent internal model controller C for the first-order time delay model (8) IMC (s), model (8) can be factorized into G m (s)=G m0 (s)e -τs , where G m0 It is a first-order minimum phase system; ideally, the internal model controller C IMC (s) should be G m0 (s) is the inverse of the closed-loop system G p (s)=G m (s) satisfies Y(s) / T set (s) = 1, thus achieving perfect tracking of the reference signal; however, due to Is not true, that is, the controller It is physically impossible to realize; Therefore, here in C IMC (s) further introduces a first-order filter Q(s) = 1 / (λs+1), where λ is the filter time constant, thus converting the controller C IMC (s) is converted into a half-true model; In summary, the controller C IMC The final form of (s) is: Step S44: Combining (13) and (14), the specific expected form of the controller C(s) is: Step S45: Combining the original form (11) and the desired form (15) of the controller C(s), derive the proportional coefficient k P and the integral coefficient k I Self-tuning scheme: Among them, the model parameter K m , H m and τ should be based on the system identification results and the water flow rate measurement value W f Real-time calculation, will follow the water flow rate W f changes with fluctuations.
6. The nonlinear internal model control method based on Smith Predictor for a wall-mounted boiler according to claim 5, characterized in that: The specific steps of step S5 are: Step S51: Since the transfer function of the control structure based on the Smith Predictor is as follows: Therefore, the closed-loop system satisfies: Where D(s) represents the form of d(t) in the s domain; Step S52: Set G p The delay-free part in (s) is G p0 (s)(i.e. G m0 (s) corresponds to the real model), the real delay of the system is τ p (i.e., the true value of the total delay τ), substituting (15) and (17) into (18), we can obtain: From formulas (19) and (20), it can be seen that if G m ≈G p , then the closed-loop system can be decomposed into a delay-free part CG p0 / [1+C(G m0 -G m +G p )] and a pure delay part Because when So from (19) and (20) we can infer that: That is, when s→0, the output Y(s) will reach T set (s), and the effect of interference D(s) on output Y(s) is 0; Step S53: Set the system nonlinearity f inv The model error of and f is considered as part of the model uncertainty and external disturbance. Then, combined with the final value theorem, it can be seen from (19) that if the input signal T set (t) and d(t) are bounded and stable even in the presence of time delay, model uncertainty, external disturbances, and f The parameter changes caused by the change, as t→∞, the system output y(t)=T o (t) will still accurately reach the set value T set , and the influence of interference on the output will be completely eliminated.
7. The nonlinear internal model control method based on Smith Predictor for a wall-mounted boiler according to claim 6, characterized in that: The specific steps of step S6 are: Step S61: When the water flow rate remains unchanged, v(t)=P b (t) = f[i gv (t)] to T o The transfer function of (t) can be summarized as: Among them, K m = β / α = η / (ζρc e W f ) and H m = 1 / α = m / (ρW f ); Step S62: Based on the test data, use Matlab's CurveFitting toolbox to perform fitting to obtain the identification results of each part of the system model: τ≈2.04s.
Citation Information
Patent Citations
Thermal generator set coordination control system based on internal model control structure
CN111429010A
Cited By
Radiometer absorption cavity temperature bistable control method based on adaptive Smith estimation
CN121364008A
